Oracle数据库表空间级的异地不完全恢复

Oracle 资料库做异地恢复时，全库的完全恢复需要很大的空间和很长的时间。 如果采用基于模式的恢复方式仅仅恢复某一模式下所有物

Oracle 资料库做异地恢复时，全库的完全恢复需要很大的空间和很长的时间。
如果采用基于模式的恢复方式仅仅恢复某一模式下所有物件，而不用去恢复所有的资料，这将节省大量的时间和空间。
这个在 Oracle 9i 版本中经常使用，在 10g 、 11g 中也可以使用一下。
下面案例是在 10g 的环境下将一个 ASM 上的 RAC 资料库的资料恢复到一个档案系统上单实例资料库中，并且是只恢复几个使用者下的资料。
操作步骤如下：
1 、将备份组拷贝到要恢复的机器上，如果空间够的话，可以将资料备份、控制档案备份和归档备份一起拷贝过来，如果不够的话，先拷贝控制档案备份和资料档案备份。
2 、关闭资料库，启动到 nomount 状态进行控制档的恢复
shutdown immediate;
startup nomount;
rman target /
restore controlfile from '/data/urpdb/urpdb_ctl_ikn3tot4_1_1.20120222';
3 、修改控制档
查询要恢复的使用者的表空间资讯
select owner,tablespace_name,count(*) from dba_segments group by owner,tablespace_name;
查找这些表空间和系统表空间（ 'SYSTEM','SYSAUX', 'UNDOTBS1','UNDOTBS2', 'UNDOTBS3', 'UNDOTBS4' ）对应的资料档案
select *
from dba_data_files
where tablespace_name in
('TSP_URP', 'TSP_URP_INDEX', 'TSP_JC', 'TSP_JC_INDEX', 'TSP_GXSJ',
'TSP_GXSJ_INDE', 'TSP_OWB', 'TSP_YJSSJCK', 'USERS', 'SYSTEM',
'SYSAUX', 'UNDOTBS1','UNDOTBS2', 'UNDOTBS3', 'UNDOTBS4');
SQL> alter database backup controlfile to trace;
修改控制档并执行：
CREATE CONTROLFILE REUSE DATABASE "URPDB" RESETLOGS FORCE LOGGING ARCHIVELOG
MAXLOGFILES 192
MAXLOGMEMBERS 3
MAXDATAFILES 1024
MAXINSTANCES 32
MAXLOGHISTORY 9088
LOGFILE
GROUP 1 (
'/opt/app/oracle/oradata/URPDB/group1_2.log',
'/opt/app/oracle/oradata/URPDB/group1_1.log'
) SIZE 10M,
GROUP 2 (
'/opt/app/oracle/oradata/URPDB/group2_2.log',
'/opt/app/oracle/oradata/URPDB/group2_1.log'
) SIZE 10M,
GROUP 3 (
'/opt/app/oracle/oradata/URPDB/group3_2.log',
'/opt/app/oracle/oradata/URPDB/group3_1.log'
) SIZE 10M,
GROUP 11 '/opt/app/oracle/oradata/URPDB/group11_1.log' SIZE 100M,
GROUP 12 '/opt/app/oracle/oradata/URPDB/group12_1.log' SIZE 100M,
GROUP 13 '/opt/app/oracle/oradata/URPDB/group13_1.log' SIZE 100M,
GROUP 14 '/opt/app/oracle/oradata/URPDB/group14_1.log' SIZE 100M,
GROUP 15 '/opt/app/oracle/oradata/URPDB/group15_1.log' SIZE 100M
-- STANDBY LOGFILE
DATAFILE
'/opt/app/oracle/oradata/URPDB/system01.dbf',
'/opt/app/oracle/oradata/URPDB/undotbs1.dbf',
'/opt/app/oracle/oradata/URPDB/sysaux.dbf',
'/opt/app/oracle/oradata/URPDB/users.dbf',
'/opt/app/oracle/oradata/URPDB/undotbs2.dbf',
'/opt/app/oracle/oradata/URPDB/tsp_urp.dbf',
'/opt/app/oracle/oradata/URPDB/tsp_urp_index.dbf',
'/opt/app/oracle/oradata/URPDB/tsp_gxsj.dbf',
'/opt/app/oracle/oradata/URPDB/tsp_urp.dbf_1.ora',
'/opt/app/oracle/oradata/URPDB/tsp_owb.ora',
'/opt/app/oracle/oradata/URPDB/tsp_jc.dbf',
'/opt/app/oracle/oradata/URPDB/tsp_jc_index.dbf',
'/opt/app/oracle/oradata/URPDB/tsp_yjssjck.dbf',
'/opt/app/oracle/oradata/URPDB/undotbs03.dbf',
'/opt/app/oracle/oradata/URPDB/undotbs04.dbf'
CHARACTER SET AL32UTF8
;
mount 资料库；
SQL> alter database mount;

select name from v$datafile;
--select 'alter database rename file '''||name||''' to '''|| replace(name,'+URPDBDG','/opt/app/oracle/oradata/URPDB')||''';' from v$datafile;
3 、将备份资料档案拷贝到测试机
rman>catalog start with '/data/urpdb/'; 执行一下，就是将这个目录下的资料档案的备份档案写入了 catalog 。
4 、 restore 资料库
select file#,name from v$datafile ； --- 看看控制档中资料档案存放的位置，，这里是 “ +URPDBDG/..." ，替换为目标资料库档存放的位置。
set linesize 300
column name format a200
set pagesize 100

select 'set newname for datafile '||file#||' to '''||replace(name,'+URPDBDG','/opt/app/oracle/oradata/urpdb')||''';' newname from v$datafile -- 如果一个目标目录不够，可以用多个目标目录
union all
select 'restore datafile '||file#||';' newname from v$datafile;

select *
from dba_data_files
where tablespace_name in
('TSP_URP', 'TSP_URP_INDEX', 'TSP_JC', 'TSP_JC_INDEX', 'TSP_GXSJ',
'TSP_GXSJ_INDE', 'TSP_OWB', 'TSP_YJSSJCK', 'USERS', 'SYSTEM',
'SYSAUX', 'UNDOTBS1','UNDOTBS2', 'UNDOTBS3', 'UNDOTBS4');

rman>
run
{
allocate channel ch01 type disk;
allocate channel ch02 type disk;
set until time "to_date('2012-02-23 10:00:00','yyyy-mm-dd hh24:mi:ss')";
set newname for datafile 1 to '/opt/app/oracle/oradata/URPDB/system01.dbf';
set newname for datafile 2 to '/opt/app/oracle/oradata/URPDB/undotbs1.dbf';
set newname for datafile 3 to '/opt/app/oracle/oradata/URPDB/sysaux.dbf';
set newname for datafile 4 to '/opt/app/oracle/oradata/URPDB/users.dbf';
set newname for datafile 6 to '/opt/app/oracle/oradata/URPDB/undotbs2.dbf';
set newname for datafile 9 to '/opt/app/oracle/oradata/URPDB/tsp_urp.dbf';
set newname for datafile 10 to '/opt/app/oracle/oradata/URPDB/tsp_urp_index.dbf';
set newname for datafile 14 to '/opt/app/oracle/oradata/URPDB/tsp_gxsj.dbf';
set newname for datafile 16 to '/opt/app/oracle/oradata/URPDB/tsp_urp.dbf_1.ora';
set newname for datafile 18 to '/opt/app/oracle/oradata/URPDB/tsp_owb.ora';
set newname for datafile 19 to '/opt/app/oracle/oradata/URPDB/tsp_jc.dbf';
set newname for datafile 20 to '/opt/app/oracle/oradata/URPDB/tsp_jc_index.dbf';
set newname for datafile 26 to '/opt/app/oracle/oradata/URPDB/tsp_yjssjck.dbf';
set newname for datafile 29 to '/opt/app/oracle/oradata/URPDB/undotbs03.dbf';
set newname for datafile 30 to '/opt/app/oracle/oradata/URPDB/undotbs04.dbf';
restore datafile 1;
restore datafile 2;
restore datafile 3;
restore datafile 4;
restore datafile 6;
restore datafile 9;
restore datafile 10;
restore datafile 14;
restore datafile 16;
restore datafile 18;
restore datafile 19;
restore datafile 20;
restore datafile 26;
restore datafile 29;
restore datafile 30;
switch datafile all;
release channel ch01;
release channel ch02;
}
5 、修改日志档和暂存档案的路径为目标资料库的路径（可与 restore 同时进行）
select member from v$logfile;

select name from v$tempfile;

select 'alter database rename file '''||member||''' to '''|| replace(member,'+URPDBDG','/opt/app/oracle/oradata/URPDB')||''';' from v$logfile;

alter database rename file '+YWKDG/ywk/onlinelog/group15_1.log' to '/opt/app/oracle/oradata2/YWKDG/group15_1.log';
。。。。。

select 'alter database rename file '''||name||''' to '''|| replace(name,'+URPDBDG','/opt/app/oracle/oradata2/urpdb/')||''';' from v$tempfile;

alter database rename file '+YWKDG/ywk/tempfile/tsp_zc_temp02.dbf' to '/opt/app/oracle/oradata2/YWKDG/tsp_zc_temp02.dbf';

6 、将归档档拷到目标资料库伺服器，
rman>catalog start with '/data/urpdb/'; 执行一下，就是将这个目录下的归档档的备份档案写入了 catalog 。
run
{
allocate channel ch01 type disk;
allocate channel ch02 type disk;
set until time "to_date('2012-02-23 10:00:00','yyyy-mm-dd hh24:mi:ss')";
recover database;
release channel ch01;
release channel ch02;
}
或则
SQL> select to_char(checkpoint_time,'yyyy-mm-dd hh24:mi:ss') from v$datafile;

S QL> select file#,checkpoint_change# from v$datafile order by checkpoint_change#;

FILE# CHECKPOINT_CHANGE#
---------- ------------------
14 2793109724
20 2793109725
6 2793232002
19 2793232002
18 2793258889
26 2793258889
9 2793281171
30 2793281171
1 2793283158
16 2793301522
29 2793301522
2 2793304369
10 2793304369
4 2793333238
3 2793333238

15 rows selected.

run
{
allocate channel ch01 type disk;
allocate channel ch02 type disk;
set until scn 2793333238;
recover database ；
release channel ch01;
release channel ch02;
}

如果 recover 资料库缺少档，可以到主库去查是哪个档，然后拷贝过来再 recover 一次。
RMAN> list backupset of archivelog logseq 7438 thread 4;

RMAN> crosscheck backupset;
delete expired backupset;

7 、 alter database open resetlogs;

更多Oracle相关信息见Oracle 专题页面 ?tid=12